Ayush Giri

M.S., Ph.D

Vanderbilt University Medical Center

Assistant Professor

PLCO (Learn more about this study)

PLCO-299

Sep 5, 2017

A systematic evaluation of the relationship between reproductive factors and gynecologic outcomes in women from the PLCO: endometriosis, uterine fibroids, and endometrial and ovarian cancer.

Endometriosis, uterine fibroids, endometrial cancer and ovarian cancer are distinct gynecologic conditions that may occur at varying time points in a woman’s lifespan. While the diagnosis, prognosis and treatment options for each of these conditions differ, literature suggests hormonal factors (exogenous or endogenous) play a key role in influencing these conditions. However, the extent to which hormonal influences are shared across these conditions is not clear. Several studies have suggested with varying degrees of precision that early age at menarche and nulliparity may increase risk for these conditions. On the other hand, increased body mass index, a factor that also influences the hormonal millieu is inversely associated with endometriosis, and positively associated with the other conditions. Studies also suggest that women with endometriosis may be at increased risk of endometrial cancer and epithelial ovarian cancer. Evidence of shared etiology between reproductive factors and these gynecologic conditions originate from different studies evaluated in different populations of women.

Furthermore, genetic variation may play a role in influencing these conditions, either directly, through interaction with other environmental/reproductive/lifestyle events, or indirectly, by influencing a mediator. For example, evidence suggests age at menarche is influenced by genetic variation. It is of interest to investigate whether reported associations between age at menarche and complex phenotypes are in part a function of genetic influence.

The Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial has a rich collection of variables describing reproductive factors and reproductive history, in addition to assessment of these gynecologic outcomes, self-reported at baseline for endometriosis and uterine fibroids, and incident assessment for endometrial and ovarian cancers. The PLCO dataset thus provides the unique opportunity to simultaneously assess the relationship between various reproductive factors, genetic variation and these gynecologic outcomes in the same cohort of women. Furthermore, it also allows for a temporally sound investigation of the associations between endometriosis and incident cancer outcomes, namely endometrial and ovarian cancer in a large group of women throughout the US.

1) Using baseline questionnaire data, we plan to systematically evaluate the associations between various reproductive factors (including age at menarche, oral contraceptive use, parity, hormone therapy use, tubal ligation) and all four outcomes (endometriosis, uterine fibroids, endometrial cancer and ovarian cancer). Analyses will be performed using appropriate regression modeling strategies while adjusting for potential confounders and key covariates. Where appropriate, analyses will be stratified by menopausal status, hysterectomy status and by race/ethnicity to clarify and inform associations.

2) We also plan to evaluate the association between endometriosis status reported at baseline and incident endometrial and incident ovarian cancer using Cox proportional hazard regression models while adjusting for key confounders, and risk factors related to life style and dietary variables for ovarian or endometrial cancer.

3) If genetic data are available (either as targeted array/sequence or genome-wide association study (GWAS), or sequencing data) in adequate numbers, we plan to investigate the role of genetic variation (as a main effect or interaction with reproductive/life-style/environmental factors) in relation to endometriosis, uterine fibroids, endometrial cancer and ovarian cancer. Summary statistics from these analyses may be presented by itself or as a collaborative effort with additional datasets in a consortium type setting. We will further utilize polygenic risk scores to assess the potential for causality for previously reported epidemiologic associations between reproductive/hormonal factors and these gynecologic conditions.